[Vladimir Jojic 971031.1245 MET]
I have a question for people who are interested in experimental/simulation
side of PCT ...
Did anyone try to discretize the range of the response and represent each
of these segments as a certain action, and let these actions have some
side-effects on other perceptions?
Thanks
Vladimir
[From Bill Powers (971031.0935 MST)]
Vladimir Jojic 971031.1245 MET--
I have a question for people who are interested in experimental/simulation
side of PCT ...Did anyone try to discretize the range of the response and represent each
of these segments as a certain action, and let these actions have some
side-effects on other perceptions?
Don't know what you mean. Can you give an example?
Best,
Bill P.
[From Rick Marken (971031.0930)]
Vladimir Jojic (971031.1245 MET) --
Did anyone try to discretize the range of the response and
represent each of these segments as a certain action, and let
these actions have some side-effects on other perceptions?
WHat you are saying sounds like a description of my "Hierarchy
of perception and control" experiment at:
http://home.earthlink.net/~rmarken/ControlDemo/HP.html
The response is a mouse press, so it is discrete: the response
either happens (press) or it doesn't (no press). The response
has an effect on three variables: shape of figures, sequential
position of figures and sequential order of size of figures.
If the subject controls one of these variables (perceptions)
then the effect of mouse presses (actions) on the other
variables is a side effect.
Why did you want to know this, by the way?
Best
Rick
--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken
[Vladimir Jojic 971031.1930 MET]
To both Bill and Rick ....
Rick Marken (971031.0930)
Vladimir Jojic (971031.1245 MET) --
> Did anyone try to discretize the range of the response and
> represent each of these segments as a certain action, and let
> these actions have some side-effects on other perceptions?WHat you are saying sounds like a description of my "Hierarchy
of perception and control" experiment at:http://home.earthlink.net/~rmarken/ControlDemo/HP.html
The response is a mouse press, so it is discrete: the response
either happens (press) or it doesn't (no press). The response
has an effect on three variables: shape of figures, sequential
position of figures and sequential order of size of figures.
If the subject controls one of these variables (perceptions)
then the effect of mouse presses (actions) on the other
variables is a side effect.
Ok, that's it. To make it more clear: ... after I calculated output, I
translate it (output) to one of the actions from a discrete set of
actions, depending on amount of output. Furthermore, I am interested in
having each of those actions affecting some other controlled variables ...
(example: I am downtown and hungry, but I don't have enough time to go
home, so I can get some junk food (side effects: cholesterol, stomach
ache, or something), just grab some chocolate (side effect: you can
guess)) or not eat at all, till later (side effects: I am hungry and
nervous))
Why did you want to know this, by the way?
I am thinking about application of simulation of PCT in sociology and
political science. For example, if you narrow voters actions to voting,
and controlled variables to couple of their essential interests (money in
the pocket, health insurance, strict laws, etc.), establish their
perceptions of controlled variables, and finally establish how certain
parties influence certain voters' controlled variables -- then you are
getting quite close to game theory, from PCT view point, and cooperative
control of controlled variables (as McClelland showed in his paper on
collective control of perceptions).
You can iterate elections, and simulate different parties, disturbing
voters' perceptions of controlled variables, see the side effects of
certain policies on voters' controlled variables, and predict how these
side effects would affect their chances on next elections, etc...
Please, don't start saying: "But, according to PCT, we don't ...."
I am still thinking about it, and if I thought it was a clear idea, I
would have told you about it in the first place, but since you asked, here
it is ...
Regards,
Vladimir
Best
Rick
--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken
[From Rick Marken (971103.0800)]
Vladimir Jojic (971031.1930 MET)
Let's forget the model for a second and try to figure out what
you are trying to explain. Apparently you are trying to explain
something like the following:
I am downtown and hungry, but I don't have enough time to go
home, so I can get some junk food (side effects:cholesterol,
stomach ache, or something), just grab some chocolate (side
effect: you can guess)) or not eat at all, till later (side
effects: I am hungry and nervous)
Ok. Based on this description I would say that we are dealing
with a person who is trying to control a perception of hunger.
This is typically done by producing the perception of food. You
can produce the perception of food by going home. But you are
also apparently controlling for being somewhere (say, work) at
a particular time. So a higher level system imagines that going
through the steps involved in going home would prevent you from
perceiving yourself back at work on time. If you are controlling
very strongly for having the food that's at home, you will find
yourself in a conflict; you will start moving towards home to
eat but this will increase the error in the system controllng
for being back a work on time. You might end up frozen by this
conflict, standing on the street corner "deciding" whether
to home to eat or get back to work on time.
One way to solve this conflict is by realizing that there might
be alternative food downtown near work. But this realization will
create its own conflict if the only food you can think of getting
downtown is "junk food" and you are controlling for not eating
junk food (for whatever reasons: controlling for its imagined
effect on cholesteral levels -- you can't perceive your cholesterol
level so any actual effect of the food on cholesteral level is
a true, irrelevant _side effect_ of control of food intake --
possible illness, or whatever).
Another way to solve the conflict is by postponing eating; you can
delay control of food intake and still be able to control for
hunger but you can't delay control of being at work on time and
still be at work on time. As you note, however, this solution
requires that you tolerate increased hunger and its possible side
effects. If you find these possible (imagined) consequences
of this solution to your eating/work on time conflict unacceptable,
you are still in the conflict.
So, in the story above, the "side effects" of controlling for one
perception are mainly disturbances to the perceptions controlled
by other control systems. The efforts made by these other control
systems to resist these disturbances create side effects that
are disturbances to the first set of control systems. So it looks
to me like what you are concerned about in your example above is
_conflict_ between control systems -- intra-personal conflict
because all of these control systems are in the same person.
I am thinking about application of simulation of PCT in sociology
and political science.
I think it would be good if you could get a hold of Bill's CROWD
program and see how that works. The operation of that program is
precisely relevant to your interest in the application of PCT to
sociology. If you can't get the program (due to FTP problems) at
least try to get a copy of the article describing how the program
works:
McPhail, Clark., Powers, William T., & Tucker, Charles W. (1992).
Simulating individual and collective action In temporary gatherings.
Social Science Computer Review, 10(1), 1-28.
Best
Rick
--
Richard S. Marken Phone or Fax: 310 474-0313
Life Learning Associates e-mail: rmarken@earthlink.net
http://home.earthlink.net/~rmarken